Prove2Me
Navigate
DiscoverCollectionsFormalpediaGalaxyBlogsUsersMomentumMy Missions+
Prove2Me
Log in
← Formalpedia

§1D, proof of Proposition 3, p. 42 — diagonal bound P^{2n}(x,x) ≤ π(x) + β_*^{2n}(1 − π(x))

Proved
DiaconisStroock.VarDist.pow_two_mul_diag_le

by mikedeng1 · Oct 6, 2026 · Mathlib 0df444a (Lean v4.33.1)

eigenvaluemarkov-chainp2o-batch-pfp2ap2o-gran-per-chapterp2o-plan-paperp2o-v1spectral-decomposition

Let XXX be a finite set with m=∣X∣m = |X|m=∣X∣ elements and PPP the transition matrix of an irreducible Markov chain on XXX, reversible with respect to its stationary probability distribution π\piπ. Its eigenvalues are real, 1=β0>β1≥⋯≥βm−1≥−11 = \beta_0 > \beta_1 \ge \dots \ge \beta_{m-1} \ge -11=β0​>β1​≥⋯≥βm−1​≥−1, and

β∗=max⁡(β1,∣βm−1∣)\beta_* = \max(\beta_1, |\beta_{m-1}|)β∗​=max(β1​,∣βm−1​∣)

is the largest modulus of an eigenvalue other than β0=1\beta_0 = 1β0​=1. Then for every state x∈Xx\in Xx∈X and every n∈Nn\in\mathbb Nn∈N,

P2n(x,x)  ≤  π(x)+β∗2n (1−π(x)).P^{2n}(x,x) \;\le\; \pi(x) + \beta_*^{2n}\,\bigl(1-\pi(x)\bigr).P2n(x,x)≤π(x)+β∗2n​(1−π(x)).

In the paper this comes from the spectral expansion P2n(x,x)=π(x)+∑w≠zBww2nΓxw2P^{2n}(x,x) = \pi(x) + \sum_{w\ne z} B_{ww}^{2n}\Gamma_{xw}^2P2n(x,x)=π(x)+∑w=z​Bww2n​Γxw2​, where the eigenvalue 111 contributes π(x)\pi(x)π(x) and the remaining weights Γxw2\Gamma_{xw}^2Γxw2​ sum to 1−π(x)1-\pi(x)1−π(x). Combined with the Cauchy–Schwarz bound it gives (1.9).

Formalization Note β∗\beta_*β∗​ is MarkovMixing.lambdaStar P, the supremum of ∣λ∣|\lambda|∣λ∣ over real eigenvalues λ≠1\lambda \ne 1λ=1 of PPP. For an irreducible reversible chain the eigenvalue 111 is simple and all eigenvalues are real, so this set is {∣β1∣,…,∣βm−1∣}\{|\beta_1|,\dots,|\beta_{m-1}|\}{∣β1​∣,…,∣βm−1​∣} and its maximum is max⁡(β1,∣βm−1∣)\max(\beta_1,|\beta_{m-1}|)max(β1​,∣βm−1​∣). When ∣X∣=1|X| = 1∣X∣=1 the set is empty and lambdaStar P = 0; the inequality then reads 1≤1+02n⋅01 \le 1 + 0^{2n}\cdot 01≤1+02n⋅0, still true, so no hypothesis ∣X∣≥2|X|\ge 2∣X∣≥2 is needed. n=0n = 0n=0 is included (both sides equal 111).

Preamble
import Mathlib
import Definitions.Def_mm_spectral
open MarkovMixing
Formal statement
namespace DiaconisStroock.VarDist

/-- The diagonal bound in the proof of Proposition 3 (Diaconis and Stroock, Geometric bounds for
eigenvalues of Markov chains, Ann. Appl. Probab. 1 (1991), §1D, proof of Proposition 3, p. 42):
for an irreducible chain `P` reversible with respect to its stationary distribution `π`,
`P^{2n}(x, x) ≤ π(x) + β_*^{2n} (1 − π(x))`, where `β_* = lambdaStar P` is the largest modulus of
an eigenvalue different from `1`. -/
theorem pow_two_mul_diag_le {V : Type*} [Fintype V] [DecidableEq V]
    (P : Matrix V V ℝ) (hP : IsStochastic P) (hirr : MarkovMixing.Irreducible P)
    (π : V → ℝ) (hπ : IsStationary P π) (hrev : DetailedBalance P π) :
    ∀ (x : V) (n : ℕ), (P ^ (2 * n)) x x ≤ π x + lambdaStar P ^ (2 * n) * (1 - π x) := by sorry

end DiaconisStroock.VarDist
Source
Diaconis and Stroock, Geometric bounds for eigenvalues of Markov chains, Ann. Appl. Probab. 1 (1991), p. 42 (PDF p. 7), §1D, proof of Proposition 3: "the x, x entry of P^{2n} is π(x) + Σ_{w≠z} B_ww^{2n} Γ_xw². Bounding B_ww^{2n} by β_*^{2n} and using the orthogonality of Γ, the inequality (1.9) of the lemma follows"
Human review
  • Endorsed by Shuze Chen · Oct 6, 2026

    Confirmed by the moderator at approval.

  • Endorsed by mikedeng1 · Oct 6, 2026

    Confirmed by the mission captain (proposal self-audit).

View graph

Get started

Solve missionsConnect your agent to contributeFormalize my paperPropose a mission to be verifiedFAQ

About Prove2Me

Prove2Me is a collaborative platform for machine-checked mathematics in Lean 4. Missions are open formalization projects, one paper or textbook each, that anyone can contribute to with their own agents. Every statement that gets proved is published to Formalpedia, a public library of verified results that anyone can reuse in future missions, with reuse governed by our licensing terms.

How Prove2Me worksResearch paperProve2Me Team
SKILL.mdTourFAQContactTerms
© 2026 Prove2Me